<< 19-07-2020 >>

---

00:00:02	*	junland quit (Quit: %ZNC Disconnected%)
00:01:03	FromDiscord	<Varriount> How does it know what preprocessor to call?
00:02:12	shashlick	It uses CC
00:02:16	*	junland joined #nim
00:04:18	shashlick	What exactly are you trying to wrap
00:05:41	*	couven92 joined #nim
00:05:54	*	fredrikhr quit (Disconnected by services)
00:06:00	*	couven92 is now known as fredrikhr
00:07:17	FromDiscord	<Varriount> So if `$CC` was blank?
00:07:33	FromDiscord	<Varriount> Because I'm on Windows, and that's not a standard environment variable?
00:08:34	shashlick	But default it calls gcc
00:08:39	shashlick	See getters
00:10:29	FromDiscord	<Varriount> Well, I set CC to clang, still no change. This is odd
00:12:58	shashlick	Ok what's your command line? And what's the output
00:13:08	shashlick	Is -d not showing any output
00:13:24	FromDiscord	<Varriount> .\toast -d -p ../../winim-contrib/output/UIAutomation.alt.h
00:13:44	shashlick	You need -n to generate Nim output
00:13:47	*	oddp quit (Ping timeout: 240 seconds)
00:13:59	shashlick	That should just print out the preprocessed output
00:14:05	*	oculux joined #nim
00:14:30	FromDiscord	<Varriount> Well, It's not generating preprocessed output either. Here's the command line and stack trace: https://gist.github.com/Varriount/c3f8a045d816bb09fbc51e3a8a1bd2b9
00:15:05	shashlick	So if you aren't seeing any output with -p there's an issue with preprocessing
00:16:53	FromDiscord	<tomck> sent a code paste, see https://play.nim-lang.org/#ix=2s1L
00:20:02	FromDiscord	<Varriount> @tomck Put `: untyped` after each parameter declaration in `accessField`
00:20:20	FromDiscord	<tomck> same issue
00:22:29	FromDiscord	<tomck> ohhhh no that's mb, my template was wrong - untyped works, that's great thanks!
00:25:29	shashlick	I need to try this
00:25:34	shashlick	Where can I find that header
00:25:41	disruptek	tomck: what lang are you coming from?
00:25:43	shashlick	That stack trace is useless
00:26:00	shashlick	@varriount only suggestion is to give full path to file
00:27:09	FromDiscord	<Varriount> shashlick: Found it. I had to change the call to `sanitizePath` in line 322 of getters.nim to `sanitizePath(noQuote=true)`
00:27:49	shashlick	Goody
00:29:04	FromDiscord	<tomck> disruptek: a mixture haha, zig/terra/rust are the ones i've used with most similar metaprogramming to nim, use c/c++ a lot too ofc, and slowly crush my soul day by day w/ javascript
00:33:53	Yardanico	I don't think zig has macros at all though
00:34:02	Yardanico	it has compile-time features, yes, but not macros :P
00:34:21	Prestige	@Elegant Beef I'm attempting to get the types of arguments of a callback proc and pass them to a hashing function, basically: https://github.com/avahe-kellenberger/EasyECS/blob/macro-testing/src/EasyECS/registry.nim#L68
00:35:06	Prestige	Code doesn't work here, trying to figure out why & how to get around the problem
00:35:16	FromDiscord	<Varriount> shashlick: Interesting... I tested it on a header file with COM definitions, and the preprocessor emitted the C definitions rather than the C++ ones. Normally you have to define a specific macro for that to happen.
00:35:38	shashlick	That's cause it's running in C mode by default
00:35:42	FromDiscord	<tomck> Yardanico: not ast-level, it has constexpr if & (more importantly) a constexpr `for` loop though, which lets you loop over a comptime list & 'generate' a line of code for each value
00:35:46	shashlick	Toast doesn't know c++
00:35:54	Prestige	It's the same thing this template is doing: https://github.com/yglukhov/ecs/blob/master/ecs.nim#L73
00:36:15	shashlick	Tree sitter can parse c++ but toast doesn't know what to do with it
00:37:09	FromDiscord	<Varriount> Ah, I see. Those blocks are guarded by a `if !defined(__cplusplus)... condition`
00:38:29	disruptek	tomck: can you compare nim to those langs for us?
00:39:09	Yardanico	disruptek: use @ to ping on discord
00:39:12	Yardanico	@tomck ^ :P
00:39:31	disruptek	eh.. if you want a ping from me, you'll switch to irc.
00:39:36	Yardanico	meh
00:39:39	FromDiscord	<Rika> lol
00:39:46	Yardanico	don't be so elitist :P
00:39:58	disruptek	nothing elitist about it.
00:40:16	FromDiscord	<Rika> just disruptek being disruptek i guess
00:40:36	FromDiscord	<Varriount> shashlick: Can `symOverride` be used to skip defines that one doesn't need (like header include-once guards)?
00:42:09	FromDiscord	<tomck> sent a long message, see http://ix.io/2s1O
00:42:14	*	enthus1ast quit (Ping timeout: 256 seconds)
00:42:44	FromDiscord	<tomck> definitely check out terra if you haven't already, it's the most interesting one of the 3
00:42:53	FromDiscord	<tomck> http://terralang.org/http://terralang.org/
00:43:07	Yardanico	@tomck I'd have to disagree about rust a bit - proc macros here operate on tokens, not AST nodes, which is quite different IMO
00:43:24	FromDiscord	<Varriount> tomk: For zig, what do you mean about compile time introspection?
00:43:28	FromDiscord	<tomck> oh, what's the (practical) difference? i haven't used them too much?
00:43:38	FromDiscord	<Varriount> Or rather, how does it do it well?
00:44:08	FromDiscord	<Rika> tokens arent processed into an ast tree yet i assume
00:44:28	Yardanico	yeah, exactly, you have less information about the code you're being given to a macro
00:44:51	Yardanico	with AST like in nim you can even inspect the types/procedures - get their definitions, etc
00:44:56	Yardanico	I mean with ast macros
00:45:08	FromDiscord	<Varriount> But more flexibility I presume, if the compiler supports it.
00:45:18	FromDiscord	<tomck> @Varriount like you can run arbitrary zig code at compile time to generate constant values based off of stuff, i haven't used it in a while but you can do stuff like generate an int type with N bits, where N is the number of ordinals in an enum you defined earlier or something silly
00:45:32	Yardanico	@tomck that's possible in Nim too of course
00:45:40	Yardanico	nim has a VM in the compiler for running Nim at compile-time
00:45:47	FromDiscord	<Varriount> tomck: Ah. Nim can do that. I wouldn't call it easy necessarily.
00:45:58	FromDiscord	<tomck> you can probably do a similar thing in nim, but it feels better integrated into zig, can label stuff as 'comptime' to enforce that it runs at compile time
00:45:58	Yardanico	nim doesn't have arbitrary-size integers, but you can do a lot of stuff with "static" too
00:46:05	Yardanico	@tomck "const" in nim :)
00:46:12	Yardanico	and {.compileTime.} for procs
00:46:12	FromDiscord	<Varriount> It's not that the generation is hard, it's analyzing the AST
00:46:28	FromDiscord	<tomck> oh cool, i didn't know that
00:46:43	FromDiscord	<Rika> const a = static: if you need a block of code to run statically w/ a return type i assume...
00:46:51	Yardanico	most of the stdlib works at compile-time too
00:46:55	FromDiscord	<tomck> rust proc macros seem to be able to do pretty bonkers things, i've only ever used them though, they're a huge pain to write
00:47:00	Yardanico	e.g. the `json` module
00:47:12	FromDiscord	<tomck> ahhhhhhhhhhh interesting, that's super cookl
00:47:18	FromDiscord	<Rika> nim macros does a fucktonne of things
00:47:27	FromDiscord	<tomck> what happens when nim transitions to arc? does the stdlib break?
00:47:31	FromDiscord	<Rika> someone tried to a macro only discord api wrapper in nim afaik lol
00:47:32	FromDiscord	<Rika> no
00:47:33	Yardanico	@tomck well, nim macros can do literally anything as long as they're given code with valid nim syntax :P
00:47:37	disruptek	tomck: interesting. looks like terra runs in a standard luajit binary, is that right?
00:47:37	FromDiscord	<Rika> arc should be plug and play
00:47:53	Yardanico	@tomck orc will be the default, and stdlib will work just fine
00:47:59	FromDiscord	<tomck> b/c i'd use nim for games if it didn't have a gc
00:48:08	disruptek	well, it doesn't.
00:48:09	FromDiscord	<tomck> will stdlib all work normally with arc?
00:48:11	disruptek	--gc:none
00:48:14	Yardanico	of course @tomck
00:48:22	Yardanico	ORC will become the default and you could always use ARC if you are sure you don't have cycles
00:48:22	FromDiscord	<tomck> well yeah, but then you're missing half the lang
00:48:28	disruptek	how do you figure?
00:48:46	Yardanico	even the default nim GC isn't really bad for games or other stuff IMO
00:48:50	FromDiscord	<tomck> like how rust has 'unsafe', but b/c all pointers are basically 'restrict' you hit UB way too quick, so yeah you *can* write unsafe code but not really like in c/c++
00:48:52	Yardanico	it's been tested and optimized for years as well :)
00:49:33	FromDiscord	<tomck> yeah nim seems mature enough, don't run into compiler segfaults like you do in zig lol
00:49:40	Yardanico	well they are there :P
00:49:44	FromDiscord	<Rika> ~~we still run into them~~
00:49:44	disruptek	the default gc isn't a stop-the-world variety, so it's generally workable for games. but arc offers even better latency.
00:49:46	FromDiscord	<Rika> just not really often
00:49:46	FromDiscord	<tomck> disruptek: it's not a standard luajit binary, so you can't use love2d with terra unforunately
00:49:58	FromDiscord	<tomck> (or maybe you can with a bit of faff?)
00:50:38	FromDiscord	<tomck> i have no idea about gc's, i just know that stutters in games make me feel sick
00:50:48	Yardanico	well refc doesn't really have stutters :P
00:50:50	FromDiscord	<Rika> not all gcs add stutter to a game
00:51:00	FromDiscord	<tomck> i saw nim has a soft-realtime gc, but its max pause was 2ms, which is... somewhat high for games
00:51:18	FromDiscord	<Rika> thats max settable pause
00:51:25	FromDiscord	<Rika> you *dont need to* set it that high
00:51:32	Yardanico	@tomck you can configure nim gc easily
00:51:34	FromDiscord	<tomck> I probably *wouldn't* be a problem, i just don't want to develop an entire game in nim then get fucked when i realise i have to restructure everything to 'tune the gc'
00:51:39	Yardanico	e.g. in a game loop let it run when you have free "time"
00:51:45	FromDiscord	<tomck> @Rika i thought that was the min pause?
00:52:01	FromDiscord	<Rika> you just said it was "max pause" in the message
00:52:09	Yardanico	https://nim-lang.org/docs/gc.html see soft realtime support here
00:52:31	FromDiscord	<tomck> oh i think it's 1ms, this is what i'm referencing: ↵> Tests show that a 1ms max pause time will be met in almost all cases on modern CPUs
00:52:52	FromDiscord	<tomck> which is probably fine, but... eh
00:52:53	Yardanico	there's a GC_step which should be very good for games
00:52:57	FromDiscord	<Rika> that's still 7 milliseconds of time if you're aiming for 120fps
00:53:08	Yardanico	so you can run all your code and then run the GC if you have "free" time left
00:53:15	Yardanico	in game loop
00:53:37	FromDiscord	<tomck> yardanico: yeah but it still might pause for 1ms
00:53:59	FromDiscord	<Rika> are you programming a rhythm game or something
00:53:59	disruptek	you wouldn't use --gc:refc for a game, anyway.
00:54:02	FromDiscord	<Rika> or a shooter
00:54:04	Yardanico	disruptek: you could easily
00:54:08	Yardanico	it would work nicely in most cases
00:54:09	disruptek	but you wouldn't.
00:54:13	Yardanico	@tomck well anyway we have orc
00:54:21	Yardanico	(orc is arc with a cycle collector)
00:54:29	Yardanico	orc will become the default and everyone will eventually forget about refc :P
00:54:48	FromDiscord	<tomck> no, i just hate stutters in games
00:54:52	Yardanico	i do too
00:55:01	FromDiscord	<tomck> i'd rather 30fps than a stutter
00:55:07	FromDiscord	<tomck> ...maybe
00:55:10	Yardanico	well GCs aren't really bad for games
00:55:14	FromDiscord	<Rika> 30fps is 32 milliseconds of delay afaik
00:55:24	Yardanico	e.g. there's a game called osu!, it's far from being graphics-intensive but it's a rhymth game
00:55:29	FromDiscord	<tomck> yeah i also dont want 30fps;)
00:55:32	Yardanico	and it's written in C# and still runs without stutters
00:55:40	FromDiscord	<Rika> 60fps is 16 milliseconds of delay
00:55:43	Yardanico	because stuttering in a rhymth game is a no-no
00:55:47	FromDiscord	<tomck> 2s kid's calling
00:55:56	FromDiscord	<Rika> it has been 2 seconds xd
00:56:00	FromDiscord	<Rika> (i kid)
00:57:09	Yardanico	@tomck for some info, you can read https://forum.nim-lang.org/t/6549 https://forum.nim-lang.org/t/5734 https://nim-lang.org/docs/destructors.html
00:57:58	*	lritter quit (Quit: Leaving)
01:02:07	*	disruptek is being spammed by some knucklehead trying to auth as him.
01:03:25	FromDiscord	<Rika> lol
01:03:57	FromDiscord	<tomck> disruptek: what, you are? who would do that?
01:04:24	disruptek	94.236.236.2
01:04:26	FromDiscord	<tomck> on irc? do people do that?
01:04:30	Yardanico	almost never :)
01:04:33	disruptek	i guess.
01:04:34	FromDiscord	<Rika> 1337 h4xx0r5
01:05:30	FromDiscord	<Rika> ~~regards osu, i'm surprised at the frame times that game can attain like what, 0.35ms on C#?~~
01:09:11	FromDiscord	<tomck> hmm, interesting - how
01:09:15	FromDiscord	<tomck> is it using unity?
01:09:17	Yardanico	no
01:09:19	Yardanico	custom engine
01:09:29	Yardanico	and it's being rewritten (open source) to a new version
01:09:32	Yardanico	which is still very fast
01:09:38	FromDiscord	<tomck> using smth like opengl or xna?
01:09:41	Yardanico	OpenTK
01:10:05	FromDiscord	<tomck> hmm interesting, @Rika you have a good machine?
01:10:57	FromDiscord	<tomck> gotta say when i used to play osu on my trash £100 computer it always seemed very zippy
01:11:03	FromDiscord	<Rika> yeah, kinda, ryzen 2700x or smth
01:11:32	FromDiscord	<tomck> do we have access to the current source?
01:11:34	FromDiscord	<Rika> no
01:11:41	FromDiscord	<Rika> it's closed, for stable
01:11:45	FromDiscord	<Rika> but lazer is open
01:12:29	FromDiscord	<tomck> hmm, do you think they're doing the old 'preallocate everything so the GC never hits & ignore the fact that we should've used c' trick
01:12:50	FromDiscord	<tomck> i've heard of a worrying number of people doing that just so they can use c#/java/whatever
01:12:58	disruptek	mm just isn't a hard problem, honestly.
01:12:59	FromDiscord	<tomck> can only assume it's horrible
01:13:08	FromDiscord	<Rika> well if theyre doing that then it's working pretty well
01:15:43	FromDiscord	<Elegant Beef> My favourite thing is minecraft benchmarks when talking about GC, you can never see the GC :D https://media.discordapp.net/attachments/371759389889003532/734216847930687528/unknown.png
01:16:57	FromDiscord	<Rika> lol
01:17:34	FromDiscord	<tomck> does MC do much gc tuning? you'd have to imagine so
01:17:57	FromDiscord	<Rika> probably but it'll still be obvious since java 😛
01:18:40	FromDiscord	<tomck> fair, minecraft runs like absolute horse shit, i almost can't believe it's so popular
01:18:56	FromDiscord	<tomck> guess people don't care - did it get re-written for consoles? or is it really laggy there too?
01:18:56	FromDiscord	<Elegant Beef> i mean it runs at 200+ fps in that benchmark
01:19:02	Yardanico	bedrock edition exists @tomck
01:19:05	FromDiscord	<Elegant Beef> They use bedrock
01:19:06	FromDiscord	<Elegant Beef> which is C++
01:19:17	Yardanico	they have shared MCPE/Windows 10 edition code nowaday
01:19:18	Yardanico	s
01:19:19	FromDiscord	<tomck> ohhh, and java edition is desktop only?
01:19:22	Yardanico	yes
01:19:28	Yardanico	you can run bedrock edition on Linux too btw
01:19:31	FromDiscord	<Elegant Beef> But the modding scene for bedrock is non existant so isnt popular
01:19:31	FromDiscord	<Rika> bedrock is windows only 🙂
01:19:34	Yardanico	it's not
01:19:35	FromDiscord	<Rika> wait rly?
01:19:38	FromDiscord	<Rika> interesting
01:19:40	Yardanico	Well, unofficially, yes
01:19:43	FromDiscord	<tomck> can you mod bedrock?
01:19:48	Yardanico	basically MCPE has x86 build which is still bedrock
01:19:53	Yardanico	and with a custom launcher you can play it
01:19:56	FromDiscord	<Rika> lol
01:19:57	FromDiscord	<Elegant Beef> Well the linux method is running an android emulator afaik
01:19:59	Yardanico	I tried and it works
01:20:05	Yardanico	@Elegant no
01:20:06	Yardanico	it's native
01:20:09	FromDiscord	<Elegant Beef> Ah
01:20:18	Yardanico	I even got it to connect my Microsoft account for online play and stuff
01:20:23	FromDiscord	<Elegant Beef> I seen someone say "It's running" and their solution was android launcher
01:20:29	FromDiscord	<tomck> lol, sounds like it's windows only
01:20:52	Yardanico	bedrock? android, ios, windows, linux (unoficially)
01:20:59	FromDiscord	<tomck> why does nobody do gamedev for linux in general?
01:21:03	FromDiscord	<tomck> is support just a nightmare?
01:21:07	Yardanico	because much less market share
01:21:08	FromDiscord	<Elegant Beef> Small population
01:21:19	Yardanico	"not worth the money"
01:21:19	FromDiscord	<Elegant Beef> More tech savy people, who are more likely to make reports
01:21:35	FromDiscord	<Elegant Beef> so you get `70% of the reports from 1% of the market`
01:21:37	FromDiscord	<tomck> yeah but you can do cross platform dev at seemingly no cost, or is it actually a huge cost once you get into the problems it causes?
01:21:46	Yardanico	depends on the engine/etc
01:21:49	FromDiscord	<Elegant Beef> I mean bedrock is ported everywhere
01:21:51	Yardanico	Linux is very diverse, yes
01:21:54	FromDiscord	<tomck> tbf they've probably also got ridiculous setups
01:21:56	FromDiscord	<Elegant Beef> You'd have to imagine it'd be simple to port to linux
01:22:02	Yardanico	but most linux people still just run ubuntu
01:22:06	Yardanico	or arch/manjaro
01:22:11	Yardanico	(you can check steam linux stats)
01:22:39	FromDiscord	<tomck> i'm running i3, bet you an official port would open in a tiny 32x32 window or something like that
01:22:45	Yardanico	not really
01:22:55	FromDiscord	<Elegant Beef> Nah that's never the issue
01:22:57	Yardanico	I've used tiling WMs for over a year now and native ports work just fine
01:23:00	FromDiscord	<tomck> i hear people whining about 'sound on linux is broken' when it comes to gamedev, is that the case? never usedi t
01:23:11	FromDiscord	<Elegant Beef> Eh people dislike pulse, and jack is hard
01:23:11	Yardanico	well we maybe can go to #nim-offtopic :)
01:23:14	Yardanico	#offtopic on IRC
01:23:14	FromDiscord	<Rika> not really no
01:23:16	Yardanico	on discord*
01:27:00	*	apahl quit (Ping timeout: 256 seconds)
01:27:50	*	RaycatWhoDat quit (Remote host closed the connection)
01:28:34	*	apahl joined #nim
01:39:32	ForumUpdaterBot	New thread by Tmsa04: Map change, see https://forum.nim-lang.org/t/6557
01:43:40	FromDiscord	<Rika> That looks like a bug :EyesShaking:
01:44:31	Yardanico	well it's not
01:44:38	Yardanico	maybe it is, but I think no :P
01:44:47	Yardanico	it's a side effect of most procs in options module becoming inline
01:45:03	Yardanico	closure proc is incompatible with a inline proc it seems
01:47:32	FromDiscord	<Rika> Oof
02:10:13	FromDiscord	<Rika> shashlick: okay so i looked at the plugin start code, and (was gonna put code to change here but its long so ill just PR it then i guess T_T)
02:10:24	shashlick	Cool
02:10:36	FromDiscord	<Rika> (this is with regards the binary monitor thread thing)
02:13:06	*	endragor joined #nim
02:18:05	*	muffindrake quit (Ping timeout: 272 seconds)
02:18:43	*	krux02_ quit (Remote host closed the connection)
02:19:19	*	muffindrake joined #nim
02:20:02	FromDiscord	<Rika> shashlick: i only need to run `nimble test` to test right, or is there anything else i need to do?
02:24:57	voltist	Is anyone here familiar with the workings of the FITS file format?
02:25:57	*	skrylar[m] looks up what this is
02:26:41	FromDiscord	<Vindaar> @voltist: surprise, surprise I am 😄 really gotta go to bed now
02:26:53	FromDiscord	<Vindaar> but there is a nim wrapper I think
02:27:21	FromDiscord	<Vindaar> https://github.com/ziotom78/nimcfitsio
02:27:24	voltist	Oh well, maybe we can chat about this later then :)
02:27:25	skrylar[m]	archival format for scientific data?
02:27:29	voltist	I'll have a look at that wrapper
02:27:50	skrylar[m]	seems to be very well documented too https://fits.gsfc.nasa.gov/standard40/fits_standard40aa-le.pdf
02:28:09	FromDiscord	<Vindaar> @skrylar yeah, mostly used in astronomy though
02:28:27	voltist	skrylar[m]: Yeah I'm looking at it for an astronomy use-case
02:28:57	FromDiscord	<Vindaar> in any case, good night 🙂
02:29:01	voltist	The documentation is good if the software you use to take images behaves well
02:29:22	voltist	@Vindaar You 2
02:29:49	voltist	Fun fact: Most astronomical software doesn't behave well
02:30:34	skrylar[m]	of course it doesn't :)
02:32:25	voltist	It would be nice if other members of my local astronomical society supported the use of free software, because there are some nice open source alternatives
02:32:44	skrylar[m]	have sometimes wondered why software doesn't just rely on dump formats and then give people a txt on how the dump works
02:33:36	skrylar[m]	then at least someone who cares enough to do interchange formats properly can do so separately, instead of everyone bungling interchange formats
02:35:28	*	waleee-cl quit (Quit: Connection closed for inactivity)
02:36:43	FromDiscord	<Rika> shashlick: i think you misremembered when you said -d:binary only opened a thread and closed after running
02:39:44	FromDiscord	<Rika> or maybe my code is wrong...
02:40:06	*	nikita_ joined #nim
02:42:46	FromDiscord	<Rika> ah no yeah the thread still handles loads with -d:binary
02:43:14	*	nikita` quit (Ping timeout: 256 seconds)
02:53:43	shashlick	Just nimble test yes
02:53:54	shashlick	It tests binary mode too
02:54:27	shashlick	Maybe if it's not compiled with threads, you force single thread binary mode
02:55:19	shashlick	But ya, binary mode doesn't justify using a thread
02:55:28	shashlick	Might just call that as a proc
02:56:28	FromDiscord	<Rika> well no the issue is that you're handling extra loads (after start) with the monitor proc
02:56:41	FromDiscord	<Rika> i dont know how to fix that yet
03:00:00	shashlick	But in binary mode it only loads then exits the thread
03:00:12	shashlick	So it's pointless to spawn a thread for that
03:00:38	shashlick	Anyway, you might be better served to see if the package is useful before trying to improve the binary mode thing
03:01:43	FromDiscord	<Rika> i tried removing the thread and had issues with the plugin unload-load in tmain
03:11:13	FromDiscord	<Varriount> @Rika What are you working on?
03:12:15	shashlick	Ok let me see
03:12:39	FromDiscord	<Rika> https://github.com/genotrance/plugins, trying to remove the reliance of threads when -d:binary is defined (since it's not needed)
03:14:50	FromDiscord	<Rika> right now i dont know what to do about pload (it works differently to punload which i'd say is weird)
03:16:53	FromDiscord	<Varriount> Plugins are hard
03:17:26	FromDiscord	<Varriount> Or rather, it's hard to make them seamless, when you're working with a statically compiled language.
03:17:33	FromDiscord	<Rika> yeah
03:20:22	shashlick	okay so Rika, you basically don't want threads in binary mode right?
03:20:34	shashlick	so you will prebuild your dlls
03:20:52	FromDiscord	<Rika> yes
03:21:24	FromDiscord	<Rika> but loading midway in the program should still be possible i hope?
03:21:44	FromDiscord	<Rika> waaaait
03:21:48	FromDiscord	<Rika> one moment, idea
03:22:16	shashlick	so the way this system works is that it loads the plugins on startup
03:22:21	FromDiscord	<Rika> yes
03:30:49	shashlick	Rika - it's a minor code change - pushing a branch
03:31:45	shashlick	are you okay if we build with threads but don't actually spawn a thread
03:31:53	shashlick	cause i can retain the channels
03:32:05	shashlick	without threads, will need an alternate way to communicate loading
03:32:33	FromDiscord	<Rika> ah yeah channels needs the threads switch
03:32:34	FromDiscord	<Rika> hmm
03:32:48	FromDiscord	<Rika> dunno what effect that has on my program but sure okay
03:36:11	shashlick	i just pushed a branch called binary - it still needs threads but doesn't create a monitoring thread
03:36:17	shashlick	quick and lazy
03:36:35	shashlick	might need an alternative to communicate dlls to load instead of channels
03:36:45	shashlick	to get rid of threads altogether
03:36:59	disruptek	rika: what are you working on?
03:37:20	FromDiscord	<Rika> ill look into it
03:37:28	FromDiscord	<Rika> disruptek: still the manga reading program lol
03:37:29	disruptek	you're not sure?
03:37:31	disruptek	oh.
03:37:53	disruptek	got a link?
03:38:32	FromDiscord	<Rika> not open right now, still preparing
03:38:50	shashlick	How do you intend using plugins
03:38:57	FromDiscord	<Rika> for the sources
03:39:02	FromDiscord	<slymilano> @dom96 @Varriount here's the smaller proof of concept that possibly shows a bug with Nim's asyncCheck. It works normally in windows and linux -- but fails on mac: https://github.com/sergiotapia/async-check
03:42:27	disruptek	this hax0r is persistent.
03:43:00	disruptek	why would anyone want to login to irc as me?
03:43:10	disruptek	i can't think of anyone with a worse reputation.
03:43:39	disruptek	maybe they want to rehabilitate my irc cred.
03:45:30	FromDiscord	<Rika> shashlick: okay so the change you just pushed was basically the same as mine, the issue is that when you call pload, nothing gets loaded and you're stuck at syncPlugins
03:45:42	FromDiscord	<Rika> also probably why your CI right now is hanging
03:46:10	shashlick	that's cause none of the plugins are compiled so there's nothing loaded
03:46:27	shashlick	you need to compile them like the plugin system would
03:46:53	FromDiscord	<Rika> let me try understanding this first wait sorry
03:46:53	shashlick	nim c --app:lib path/to/plg.nim
03:46:58	shashlick	no problem
03:46:59	FromDiscord	<Rika> i know
03:47:05	FromDiscord	<Rika> theyre already compiled
03:47:37	FromDiscord	<Rika> ok so loadPlugins is basically "recompile and load again"?
03:48:54	FromDiscord	<Rika> i dont get it wait
03:49:17	*	audiophile_ joined #nim
03:54:13	FromDiscord	<Rika> ill do something else for now, maybe ill understand if i leave it for now
03:55:04	shashlick	i see that the plugins load and run for some time then just hang - will debug that
03:56:15	*	apahl quit (Ping timeout: 272 seconds)
03:56:47	*	apahl joined #nim
03:57:35	shashlick	ah yes - pload won't do anything since there's no background thread anymore
03:58:17	shashlick	so binary mode still needs the background thread to load/unload stuff
03:58:31	shashlick	it just doesn't monitor nim file for changes or compiles them
03:58:50	shashlick	changing binary mode to run everything in a single thread might be non-trivial
04:01:36	FromDiscord	<Varriount> @slymilano Thanks!
04:06:02	*	supakeen quit (Quit: WeeChat 2.8)
04:06:31	FromDiscord	<slymilano> hope it helps, and double hope i'm the one with a typo somewhere lol
04:06:38	*	supakeen joined #nim
04:11:06	Zevv	Hey mratsim: https://medium.com/swlh/optimizing-ray-tracing-in-haskell-3dc412fff20a
04:24:41	*	Zectbumo quit (Remote host closed the connection)
04:26:09	FromDiscord	<Rika> shashlick: exactly! thats what i mean
04:28:01	shashlick	okay cool
04:28:31	*	theelous3 quit (Read error: Connection reset by peer)
04:28:42	shashlick	do you expect to unload/load plugins
04:28:49	shashlick	or just load in the beginning
04:29:46	FromDiscord	<Rika> well, for now, i assume it's only on the beginning so at the binary branch's current state its fine
04:30:03	FromDiscord	<Rika> but maybe in the future i will want to load and unload? but that's the future
04:30:09	FromDiscord	<Rika> i'll pester you again when i do i guess
04:30:14	FromDiscord	<Rika> for now, this is fine i suppose
04:30:34	shashlick	okay I'll think about it and see if there's a better way to do this single threaded
04:30:45	shashlick	but i still don't get what's the issue with the background thread in your use case
04:32:11	shashlick	is there a way to say `xyz = abc` and then wherever xyz shows up, Nim replaces it with abc
04:33:39	shashlick	like you can in C with `#define xyz abc`
04:45:15	FromDiscord	<Rika> well my concern may be misplaced
04:45:31	FromDiscord	<Rika> so maybe dont invest too much time into this because i might be wrong with my concern
04:45:47	FromDiscord	<Rika> which is that would this be able to run in a single core computer?
04:46:43	shashlick	ya sure why not - the OS should figure out which thread gets to run for how long
04:48:06	FromDiscord	<Rika> i see
04:49:13	shashlick	and the monitor thread sleeps for the most part so it gives up control
04:49:55	FromDiscord	<Rika> okay so there isnt much issue with spawning threads other than "it takes long to spawn one and tear one down"?
04:50:42	shashlick	it can be problematic if its done a lot but in this specific case, the monitor thread is spawned at startup and just sits around
04:51:07	FromDiscord	<Rika> yeah okay thanks so maybe i shouldnt worry about this then
04:51:57	shashlick	i'm open to feedback, see how it goes and let me know
04:52:25	shashlick	but i'm much more curious to see if it provides capabilities that make it worth it
04:52:39	shashlick	i know it was super useful for feud
04:53:11	*	audiophile_ quit (Quit: Default Quit Message)
04:54:32	FromDiscord	<Rika> i dont see how it wouldnt; well maybe it might die trying to send megabytes of data from one plugin to another but oh well
04:55:18	shashlick	everything is in the same memory space so there really shouldn't need to be much transferring going on
04:55:25	shashlick	at least with boehm, no idea what arc requires yet
04:57:32	FromDiscord	<Rika> well my program is still far from done so it might take a few months til i give any useful feedback
04:58:06	disruptek	months?
04:58:24	FromDiscord	<Rika> i'm slow
04:58:27	FromDiscord	<Rika> this program is massive
04:58:35	disruptek	massive?
04:58:44	FromDiscord	<Rika> lol
04:59:09	shashlick	disruptek is too prolific to relate to us mere mortals
04:59:34	FromDiscord	<Rika> disruptek is super quick at coding, i mean look at where incre--
04:59:37	disruptek	i've been stuck on the same bug all day.
05:00:19	disruptek	how big is `massive`, though?
05:02:33	FromDiscord	<Rika> idk really, i'm still thinking out the plugin api for my program and all i've finished is the model
05:04:08	*	marnix joined #nim
05:04:43	shashlick	i've been pseudo stuck as well
05:04:50	shashlick	stuff just takes so long to get right
05:06:46	FromDiscord	<Rika> sometimes i wonder why i'm a programmer 🙃
05:12:09	disruptek	i wonder that about you daily.
05:13:01	FromDiscord	<Rika> oof lol
05:16:15	*	narimiran joined #nim
05:28:45	*	disruptek <Unknown user (via SASL):94.236.236.2> failed to login to disruptek. There have been 39 failed login attempts since your last successful login.
05:29:05	disruptek	holy shit, i know why this is broken.
05:29:24	FromDiscord	<Rika> ?
05:29:38	*	marnix quit (Ping timeout: 260 seconds)
05:30:08	disruptek	i'm copying a proc inside a pragma macro... that is on the same proc.
05:30:28	disruptek	merely applying the ast is enough to trigger its evaluation.
05:30:52	FromDiscord	<Rika> what are you making?
05:30:59	disruptek	bugs.
05:31:28	FromDiscord	<Rika> same
05:32:34	disruptek	i'm working on cps.
05:32:56	disruptek	also trying to guess my irc password.
05:43:33	disruptek	i guess that's not the problem. man, it's nuts.
05:45:52	*	Zectbumo joined #nim
05:46:21	shashlick	It will be nice to work on something easy for a change
05:47:36	shashlick	I don't know what convinced me that coding is a fun hobby
05:48:02	disruptek	what's easy?
05:49:09	FromDiscord	<Rika> hello world
05:50:00	disruptek	i'm thinking of making a sailboat with an ai captain.
05:53:43	Zevv	disruptek: still chewing on the cps?
05:53:54	disruptek	yeah.
05:53:57	Zevv	what's new
05:54:13	Zevv	does any of my initial stuff hold, or is it all nonsense
05:54:23	disruptek	it's bitching at me that i'm redefining a proc and i cannot for the life of me understand how.
05:54:54	Zevv	ooh I recall having the same thing. I had to do a nasty workaround because I also didn't understand any of that
05:55:03	disruptek	i couldn't find your spam about what wouldn't work.
05:55:14	Zevv	but I can't remember the details, sorry
05:56:08	Zevv	I'm not totally sure what would or would not work. There's only these papers to go on to start with, which only has some trivial control flow as examples
05:56:30	Zevv	The implementation of the C version is not something I can read I'm afraid
05:56:44	disruptek	i couldn't find the code.
05:57:03	Zevv	it's somewhere on github iirc, lemme check
05:57:18	icy	@treeform, so i'm using ws to communicate with an xmpp server, so i'm setting `protocol = "xmpp"` when i initialize a websocket. but i get a type mismatch error: https://termbin.com/1yge
05:58:16	Zevv	disruptek: https://github.com/kerneis/cpc
05:58:32	Zevv	it never took off, 7 years of mold and dust it seems
05:58:41	disruptek	thanks.
05:59:53	disruptek	well, i have to crash. i hate to leave this bug but i'm not making progress anymore.
06:00:02	Zevv	nightynight
06:00:05	disruptek	peace.
06:00:20	Zevv	hearts and minds disruptek, hearts and minds!
06:00:27	disruptek	ikr
06:04:07	*	Vladar joined #nim
06:09:29	FromDiscord	<Varriount> So, it turns out the program I've spent the last 4 days on has been made obsolete by Nimteropt
06:10:22	FromDiscord	<Varriount> Or mostly obsolete, anyway.
06:11:06	Zevv	so something new was added to nimterop today, or did you just work on stuff that nimterop already could do
06:11:53	FromDiscord	<Varriount> The latter
06:12:07	Zevv	well, that's your own fault then, isn't it :)
06:13:20	FromDiscord	<Rika> lol
06:13:47	FromDiscord	<Varriount> The only functionality it has that Nimteropt doesn't is the ability to perform regex replacements on the input files before/after preprocessing
06:14:00	*	solitudesf joined #nim
06:14:11	Zevv	but luckily we can do that with a single line of perl!
06:14:58	FromDiscord	<Varriount> And we can generate that line by bashing our heads against keyboards repeatedly.
06:15:08	Zevv	sure thing :)
06:15:15	FromDiscord	<Rika> given a long amount of time
06:15:16	FromDiscord	<Rika> sure
06:15:52	Zevv	I use perl a lot. `perl -i -pe 's/this/that/g' *` is the only thing I ever use it for though
06:16:22	FromDiscord	<Varriount> Same here. It's a more portable version of sed
06:16:29	Zevv	nothing is more portable then sed
06:16:34	FromDiscord	<Rika> more portable? why?
06:16:43	FromDiscord	<Varriount> Not if you're on OSX or BSD
06:17:08	FromDiscord	<Varriount> They have slightly different versions of sed
06:17:10	Zevv	no sed on osx or bsd? no way!
06:17:56	FromDiscord	<Varriount> They have sed, however it's a different version or implementation.
06:18:03	Zevv	oh, right
06:18:20	FromDiscord	<Varriount> If I recall correctly, OSX sed lacks a -i option
06:18:31	Zevv	which is the whole point
06:19:02	FromDiscord	<Rika> imagine if different operating systems also had slightly different versions of programming languages...
06:19:30	FromDiscord	<Rika> "oh wait"
06:26:08	*	NimBot joined #nim
06:37:03	*	marnix joined #nim
06:39:47	*	fredrikhr quit (Ping timeout: 240 seconds)
06:49:23	*	arecaceae quit (Read error: Connection reset by peer)
06:50:38	*	arecaceae joined #nim
06:50:50	*	fredrikhr joined #nim
06:54:01	shashlick	@varriount that's why I recommend the api instead of the command line
06:54:19	shashlick	Sky's the limit with pure Nim
06:55:51	shashlick	And now that you have the ability to generate standalone wrappers, it's a no brainer to use it instead of cli plus bash or something
06:56:30	shashlick	Look at cimport and build for details
06:58:01	shashlick	https://nimterop.github.io/nimterop/cimport.html
07:15:01	FromDiscord	<Rika> this may be dumb, but how do you get the nimble version in the package's nim code
07:17:26	*	nikita_ is now known as nikita`
07:21:12	FromDiscord	<Varriount> shashlick: But cImport is static, right? If I wanted to make an executable utility, it would have to work at runtime.
07:24:16	FromDiscord	<Varriount> And forcing every user of a library to generate those files isn't feasible for this use-case. Generating that output file I showed you take about 10-20 seconds (so imagine importing 5 of them)
07:24:52	FromDiscord	<Varriount> (edit) 'them)' => 'them - it would be over a minute!)'
07:29:33	*	FromDiscord quit (Remote host closed the connection)
07:29:49	*	FromDiscord joined #nim
07:35:07	FromDiscord	<Elegant Beef> How well does nimterop work with C macros?
07:36:25	FromDiscord	<Varriount> @Elegant Beef It utilizes the C preprocessor to expand C code, although I don't know if it translates macros to templates.
07:36:57	*	idf31 joined #nim
07:36:58	FromDiscord	<Elegant Beef> So that sounds like "Pretty good"? 😄
07:57:34	*	marnix quit (Read error: Connection reset by peer)
07:57:44	*	marnix joined #nim
08:02:15	*	marnix quit (Ping timeout: 258 seconds)
08:03:15	*	marnix joined #nim
08:15:15	*	Zectbumo quit (Remote host closed the connection)
08:18:05	*	marnix quit (Read error: Connection reset by peer)
08:26:16	*	marnix joined #nim
08:32:59	*	marnix quit (Ping timeout: 240 seconds)
08:34:54	*	nikita` quit (Quit: leaving)
08:41:09	*	nikita` joined #nim
08:51:20	*	oddp joined #nim
08:55:52	*	couven92 joined #nim
09:11:19	*	fredrikhr quit (Quit: Client disconnecting)
09:11:29	*	couven92 is now known as fredrikhr
09:11:41	*	marnix joined #nim
09:12:19	*	enthus1ast joined #nim
09:21:55	*	bung joined #nim
09:38:10	*	arecaceae quit (Remote host closed the connection)
09:38:34	*	arecaceae joined #nim
09:39:32	*	Senketsu joined #nim
09:49:38	*	synthmeat quit (Quit: WeeChat 2.8)
09:51:00	*	synthmeat joined #nim
09:55:21	Oddmonger	proc modif_seq( pseq: ref seq[int], i:int, v:int): pseq[i]=v # inim says : « Error: type expected »
09:56:08	Oddmonger	ah
09:56:26	Oddmonger	i see, confusion between ':' and '='
09:56:44	Oddmonger	as ':' is used for loops :/
09:59:40	planetis[m]	Prestige: so what you are trying to create is an AnyMap data structure, right?
10:00:13	*	krux02 joined #nim
10:01:10	planetis[m]	its kind of a hard topic, I tried too, but at the end I just made a World object just holding seqs of components (of different types)
10:01:48	planetis[m]	obviously not suitable for a game engine/framework
10:02:52	planetis[m]	also since there is a timezone difference with most of us, could you please post in the forum?
10:13:59	FromDiscord	<lqdev> @planetis what's an AnyMap?
10:14:00	*	marnix quit (Read error: Connection reset by peer)
10:14:26	*	marnix joined #nim
10:17:23	*	narimiran quit (Ping timeout: 240 seconds)
10:48:04	planetis[m]	stores one value for each type
10:49:11	planetis[m]	so you can do ``map[Location]`` and returns a ``seq[Location]``
10:50:20	planetis[m]	its dynamic programming (since it hides types) but useful for an ecs framework
10:51:38	FromDiscord	<lqdev> also, what's bad about a World holding seqs of components? feels like a very simple solution that does the job
10:52:58	*	leorize quit (Remote host closed the connection)
10:53:40	*	leorize joined #nim
10:53:49	planetis[m]	sure it is I personally prefer it, but how are you going to make an ecs framework?
10:54:04	FromDiscord	<lqdev> through macros, probably
10:55:32	planetis[m]	hm, like that I guess: https://github.com/rlipsc/polymorph
10:56:11	FromDiscord	<lqdev> yeah
10:56:17	FromDiscord	<lqdev> that's one of my inspirations
10:56:40	planetis[m]	it's kind of complex, don't you think?
10:57:22	FromDiscord	<lqdev> I think I can make it simpler
10:57:35	planetis[m]	cool
10:58:21	planetis[m]	...how?
11:00:18	*	Senketsu quit (Read error: Connection reset by peer)
11:00:23	FromDiscord	<lqdev> you'll have to wait and see.
11:00:43	FromDiscord	<lqdev> don't really know how to describe it.
11:00:56	FromDiscord	<lqdev> one thing i don't like about polymorph is that it seems to use global state.
11:04:47	*	nikita` quit (Quit: leaving)
11:06:10	*	Senketsu joined #nim
11:11:46	*	Senketsu quit (Read error: Connection reset by peer)
11:15:14	planetis[m]	maybe a ``registerComponent`` macro that accepts type section and another ``makeWorld`` called in a world variable declaration after the colon
11:19:43	*	fanta1 joined #nim
11:27:59	*	waleee-cl joined #nim
11:34:01	FromDiscord	<lqdev> sent a code paste, see https://play.nim-lang.org/#ix=2s3F
11:34:04	FromDiscord	<lqdev> @planetis ↑
11:36:10	FromDiscord	<lqdev> then you can trigger all of the system interface procs individually, eg. `world.draw(target, graphics, step)`
11:36:19	FromDiscord	<lqdev> and they execute for each entity
11:36:38	FromGitter	<sealmove> where was that static version of walkDir? I forget.
11:37:05	FromDiscord	<lqdev> of course there'll also be `world.newEntity(Transform(position: vec2f(1.0, 1.0)))`
11:37:20	FromDiscord	<lqdev> sealmove: does walkDir not work on compile time?
11:37:30	FromGitter	<sealmove> it doesn't
11:37:55	FromGitter	<sealmove> but I remember there is a proc in some sub-module of stdlib that does work
11:40:24	*	hoijui joined #nim
11:45:28	Oddmonger	i don't understand how to use ref for a seq (but i guess it will the same problem with everything): https://play.nim-lang.org/#ix=2s3G
11:48:23	FromGitter	<sealmove> I don't think `ref` is an operator
11:49:29	FromGitter	<sealmove> also if you want a mutable parameter you use `var`, not `ref` in the parameter list
11:49:57	FromDiscord	<lqdev> Oddmonger: https://play.nim-lang.org/#ix=2s3H
11:52:08	Oddmonger	oh
11:52:39	Oddmonger	thank you, i had understood that "var" and "ref" as equivalent, for reference
11:52:54	FromDiscord	<lqdev> `ref` means ref-counted
11:52:59	FromDiscord	<lqdev> not reference
11:53:42	Oddmonger	/o\
11:57:11	FromGitter	<sealmove> hmm this is weird, the module I was talking about is "oswalkdir". Apparently it's deprecated and functionality is moved to "os". The thing is... you can import oswalkdir and use its procs at static context, but not you can't use the same procs from os module.
11:57:31	planetis[m]	lqdev: I coded my version in parallel :P https://play.nim-lang.org/#ix=2s3J
11:59:15	planetis[m]	hi sealmove!
11:59:36	FromGitter	<sealmove> hi!
12:04:32	planetis[m]	macrocache is already a thing? awesome https://nim-lang.org/docs/macrocache.html
12:06:01	*	supakeen quit (Quit: WeeChat 2.8)
12:06:43	*	supakeen joined #nim
12:13:19	FromDiscord	<tomck> How can i get the address of a `ref T`? Dumb question but i just can't find it anywhere
12:14:18	FromDiscord	<tomck> as in, convert `ref T` to `ptr T`
12:16:48	FromGitter	<sealmove> did you try `addr` operator?
12:18:12	solitudesf	you can just cast it
12:25:22	FromDiscord	<tomck> ohhh its `x[].addr`
12:25:31	FromDiscord	<tomck> at least, that seems to work
12:25:43	FromDiscord	<tomck> (is that correct?)
12:26:09	*	letto_ joined #nim
12:27:30	*	idf31 quit (Remote host closed the connection)
12:27:31	solitudesf	yep, that works too
12:28:14	*	zedeus_ joined #nim
12:28:40	*	zama_ joined #nim
12:29:05	*	arecacea1 joined #nim
12:29:48	FromDiscord	<tomck> brill ta
12:32:22	FromGitter	<sealmove> seems correct
12:32:52	*	Cthalupa- joined #nim
12:32:55	*	arecaceae quit (*.net *.split)
12:32:56	*	Cthalupa quit (*.net *.split)
12:32:56	*	letto quit (*.net *.split)
12:32:56	*	casaca quit (*.net *.split)
12:32:56	*	bacterio quit (*.net *.split)
12:32:56	*	nisstyre quit (*.net *.split)
12:33:03	*	zedeus quit (*.net *.split)
12:33:03	*	zama quit (*.net *.split)
12:33:03	*	zama_ is now known as zama
12:40:16	*	nisstyre joined #nim
12:40:17	*	bacterio joined #nim
13:01:37	planetis[m]	Is it a good idea to have accessors with typedesc like ``proc `[]`(a: AnyMap, t: typedesc[Position]): seq[Position]`` or make the field names stable or use an enum?
13:02:42	planetis[m]	I mean in terms of hidden gotchas and compiler bugs
13:03:00	*	Senketsu joined #nim
13:04:57	FromDiscord	<Vindaar> if you mean `[]` is a generic, I use that pattern both in nimhdf5 and ggplotnim. works well. If you want to essentially provide overloading via different types, that should also work well
13:05:23	FromDiscord	<Vindaar> haven't really encountered any issues
13:06:15	*	opDispatch quit (Quit: Konversation terminated!)
13:06:23	*	bussller joined #nim
13:06:44	bussller	does nim has a run time?
13:07:05	bussller	all the gc'ed languages have run times and hence nim has too.
13:07:09	bussller	am i right?
13:08:25	*	bussller quit (Client Quit)
13:10:52	bung	you mean vm ?
13:11:45	bung	language has runtime, what language has no run time ?
13:12:23	*	marnix quit (Ping timeout: 256 seconds)
13:13:02	planetis[m]	Vindaar: `[]` is the arrays/seqs/tables accessor, i don't use generics but the macro im writing will output proc overloads for every possible type, one for get and another for mget
13:13:29	solitudesf	bung, https://en.wikipedia.org/wiki/Runtime_system
13:15:32	planetis[m]	https://play.nim-lang.org/#ix=2s3J see the commented out code
13:17:45	Oddmonger	'$' can be something else than char to string ?
13:17:52	Oddmonger	i see it there: https://pragmagic.github.io/godot-nim/v0.7.8/planes.html
13:18:08	Oddmonger	(4th proc definition)
13:19:24	planetis[m]	seems correct to me
13:19:53	FromDiscord	<demotomohiro> It seems `$` is overloaded for `Plane` type.
13:20:12	bung	solitudesf I read completed whats your point ?
13:20:16	planetis[m]	`$` return type must be a string nothing else
13:25:24	planetis[m]	I think I won't use this typedesc thing and access fields directly!? not sure if its any better bc now i will have to provide an api to overwrite the name of the generated fields
13:26:11	FromDiscord	<demotomohiro> You can define `$` that does't return string.↵`$` operator returns string is just our convention.↵https://play.nim-lang.org/#ix=2s49
13:28:59	FromDiscord	<juan_carlos> $ can be a template that just overwrites, so I think is Ok. :)
13:32:06	FromDiscord	<Vindaar> > [] is the arrays/seqs/tables accessor↵I'm aware of that, but I was wondering if that is going to be generic or not (could just be ```proc `[]`[T](a: AnyMap, t: typedesc[T]): seq[T]``` is what I mean). But for your usecase the overloading is easier I suppose. And that should work, yes
13:38:31	*	marnix joined #nim
13:39:08	*	Senketsu quit (Read error: Connection reset by peer)
13:55:06	Oddmonger	sorry, was watching the logo-godot in nim-conf
13:55:21	Oddmonger	so "$" is a valid proc name ?
13:56:39	Oddmonger	ah with ``, ok thanks for the sample
13:57:46	*	swamptest1[m] joined #nim
14:07:23	planetis[m]	one reason not to do this
14:08:01	planetis[m]	...accessors is that Nim might produce a copy of the seq
14:09:01	planetis[m]	hopefully not with the latest gc arc and lent / cursor inference
14:10:05	FromDiscord	<Vindaar> oh, yeah if you want to just provide access to the seqs with a unified interface without a copy this is of course a problem
14:10:16	FromDiscord	<Vindaar> why not just use a template?
14:11:00	planetis[m]	yes that would be a solution
14:22:20	Oddmonger	what it «is type of (float) and has to be discarded» ? https://play.nim-lang.org/#ix=2r6H
14:22:38	*	casaca joined #nim
14:22:47	Oddmonger	if i put «discard» in the function, it's ok (but won't return anything)
14:23:20	Oddmonger	correct link : https://play.nim-lang.org/#ix=2s4x
14:23:54	Oddmonger	well now it works…
14:26:09	Oddmonger	i swear it wasn't working 5 minutes ago, and i've changed nothing
14:26:19	*	endragor quit (Remote host closed the connection)
14:26:21	*	nikita` joined #nim
14:26:47	*	endragor joined #nim
14:27:06	supakeen	That paste works for you? Because it says `ecoh` :)
14:31:10	Oddmonger	yes i messed with that (1st link given was without echo)
14:31:26	*	endragor quit (Ping timeout: 260 seconds)
14:32:18	Oddmonger	(and was even worst because modified for trying to rid out of the « is type of (float) and has to be discarded »
14:32:38	Oddmonger	but nevervmind, blame the heat and the lack of sleep :)
14:40:32	disruptek	why would anyone, ever, want to be me?
14:44:01	planetis[m]	you mean why would someone be jealous of you?
14:45:22	*	endragor joined #nim
14:46:01	planetis[m]	you generally don't want / need that, trust me
14:47:23	disruptek	some idiot has tried to login to irc as me 96 times.
14:48:10	planetis[m]	lol
14:49:47	*	hoijui quit (Ping timeout: 260 seconds)
14:50:39	solitudesf	sorry, tried to delete your freenode account
14:50:53	*	endragor quit (Ping timeout: 258 seconds)
14:51:59	*	sschwarzer joined #nim
14:52:38	FromDiscord	<dom96> 96 times eh, that's a pretty specific number
14:53:31	sschwarzer	If someone of you has a Medium account, you may want to add a comment to that article, https://medium.com/better-programming/a-python-substitute-i-tried-out-the-best-programming-language-youve-never-heard-of-9e29cd1893c0 , that compiling Nim with optimizations switched on will make the runtime a lot faster. :-) I entered a Github issue on that, https://github.com/yakkomajuri/fibonacci-benchmark/issues/1
14:53:32	disbot	➥ Compile Nim code with optimization
14:54:04	sschwarzer	But most people won't look at the Github issues, only read the article.
14:54:08	FromGitter	<machineko> Hey guys its possible to do smth like this ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5f145e9018ddba6ef5e3a1fe]
14:54:35	FromGitter	<machineko> ?
14:54:52	planetis[m]	so it happened while you were identified?
14:56:19	planetis[m]	disruptek: ^
14:58:41	disruptek	~paste
14:58:42	disbot	paste: 11a frowned-upon behavior in chat; please use a service such as https://play.nim-lang.org/ or http://ix.io/ or https://gist.github.com/ and supply us a URL instead. -- disruptek
14:58:42	planetis[m]	machineko: yes use the when statement https://nim-lang.org/docs/manual.html#statements-and-expressions-when-statement
15:00:48	*	krux02 quit (Remote host closed the connection)
15:08:12	FromGitter	<machineko> @planetis thanks :)
15:09:00	*	kungtotte quit (Read error: Connection reset by peer)
15:09:17	*	kungtotte joined #nim
15:16:23	FromDiscord	<Vindaar> disruptek, what you on about? gitter code blocks are handled fine
15:18:51	*	sschwarzer quit (Quit: leaving)
15:25:34	*	marnix quit (Ping timeout: 260 seconds)
15:27:47	*	endragor joined #nim
15:29:05	*	livcd joined #nim
15:35:23	*	endragor quit (Ping timeout: 240 seconds)
15:56:49	disruptek	well, people on irc cannot read them.
15:57:13	disruptek	they render as a 3-char wide div on my screen.
15:57:49	FromDiscord	<Vindaar> it automatically inserts a link to the gitter page though
15:58:05	disruptek	yes, this is the problem exactly.
15:58:11	disruptek	the gitter page is fucking stupid.
15:59:01	FromDiscord	<Vindaar> cause you can't be arsed to open the monstrosity that is gitter for a code paste?
15:59:40	disruptek	are you saying you don't believe me that it's stupid, or are you saying that what i say is stupid is something that i don't have right in front of my eyes?
16:01:13	FromDiscord	<Vindaar> I'm saying it's stupid to complain about that. gitter might not be ix.io in terms of size, but it's a perfectly fine compromise
16:01:28	*	B4s1l3 joined #nim
16:01:29	*	B4s1l3 is now known as opDispatch
16:02:33	FromDiscord	<lqdev> @Vindaar FYI this is how the gitter page looks when opened on a horizontal split. https://media.discordapp.net/attachments/371759389889003532/734440045456654438/unknown.png
16:02:59	disruptek	https://i.imgur.com/ff4y3bU.png
16:03:38	*	endragor joined #nim
16:03:59	disruptek	i usually don't complain, but at the moment i don't have access to my css styler that fixes it.
16:04:33	FromDiscord	<lqdev> holy fuck that is such a thicc sidebar
16:04:38	FromDiscord	<lqdev> it's not that terrible on my screen
16:04:43	disruptek	it's a perfectly fine compromise for vindaar, but i'll keep my mouth shut and ignore gitter pastes going forward.
16:06:26	FromDiscord	<Vindaar> I'm not saying it's perfect, but a) you can click the menu button in the bottom right to make the user list small and b) if the convenience makes it easier for people to ask questions I'm fine with that
16:08:57	disruptek	what's nice about ix is that my bot can make sense of it trivially.
16:09:21	disruptek	but, as you say, i'm just a stupid programmer.
16:09:32	*	theelous3 joined #nim
16:09:42	disruptek	i need some macro help if there's anyone clever available.
16:10:18	FromDiscord	<Vindaar> > but, as you say, i'm just a stupid programmer.↵@disruptek dude, interpret stuff however you want, I don't care
16:11:39	*	lritter joined #nim
16:12:33	disruptek	sure. i gave you two opportunities to not call me stupid but you chose a third way. it's no big deal, but don't try to make it sound like you didn't insult me. i don't care either. 😘
16:13:41	disruptek	https://github.com/disruptek/cps
16:13:48	disruptek	the test file demonstrates.
16:14:15	disruptek	what i want to do is to have my .cps. rewrite a proc and make significant changes to an untyped proc body.
16:14:52	disruptek	i want it untyped because i want to allow cps_foo(a, b) to turn into `return cps_foo(c, a, b)` among other things.
16:15:12	disruptek	and cps_foo(a, b) is otherwise a compiler error.
16:16:57	disruptek	it feels like a compiler bug, in that something is being typed out-of-phase. but it could be that i just don't understand macros.
16:21:16	FromDiscord	<tomck> Is there any way to run tests, but not leave the binary in the tests folder? it's a pain for my git repo, hard to ignore - or at least suffix them with .bin?
16:22:13	disruptek	you can use --outdir or --out.
16:22:25	disruptek	see also the somewhat-hidden `nim --fullhelp`
16:22:26	FromDiscord	<lqdev> nim 1.4 is also going to have `nim r` which stores the binaries in nimcache
16:23:15	disruptek	also, --outdir="$nimcache" would work.
16:23:16	FromDiscord	<tomck> lol --fullhelp
16:23:54	FromDiscord	<tomck> that seems to work
16:23:59	disruptek	nice.
16:24:00	FromDiscord	<tomck> the --outdir="$nimcache"
16:24:03	FromDiscord	<tomck> ta
16:24:06	FromDiscord	<Vindaar> disruptek, what's the problem you have trying to rewrite the proc body? your macro can just spit out a proc with modified body
16:24:27	disruptek	so what should i change?
16:25:04	FromDiscord	<Vindaar> as I said, I don't know where it breaks
16:25:28	disruptek	sorry, i didn't see where you said that.
16:25:31	FromDiscord	<Vindaar> here just modify the body, no? https://github.com/disruptek/cps/blob/master/cps.nim#L537
16:26:21	FromDiscord	<Vindaar> if you don't mind an ugly hack as an example:
16:26:31	FromDiscord	<Vindaar> https://github.com/Vindaar/TimepixAnalysis/blob/master/Analysis/ingrid/private/geometry.nim#L24-L28
16:26:34	FromDiscord	<Vindaar> used here:
16:26:48	FromDiscord	<Vindaar> https://github.com/Vindaar/TimepixAnalysis/blob/master/Analysis/ingrid/private/geometry.nim#L492
16:26:56	FromDiscord	<Vindaar> but the meat is here:
16:27:05	FromDiscord	<Vindaar> https://github.com/Vindaar/TimepixAnalysis/blob/master/Tools/CompareMarlinTpa/UseMarlinRotAngle/geometry.nim#L515-L534
16:29:43	disruptek	that's a solution to one part of the problem but not the other.
16:30:18	FromDiscord	<Vindaar> and which part does it not solve?
16:31:22	disruptek	when the body turns cps_foo(a, b) into return cps_foo(c, a, b), the compiler gives a type mismatch error.
16:32:04	disruptek	it got none -- same as `proc cps_foo() =`. it wanted a proc yielding something, etc.
16:32:31	disruptek	it's like the old type is bleeding into the new ast.
16:34:34	FromDiscord	<Vindaar> I'm not quite following, but what you describes happens if you leave `nnkSym` anywhere in the ast you re-emit. make sure to replace all symbols by idents
16:34:37	disruptek	i thought i would provide two procs and choose which to use at compile-time but maybe i need to just have different implementations at compile-time.
16:35:06	disruptek	yeah, that happens if you try to re-use a `.name`, for example.
16:36:26	disruptek	hmm, i can't have different impl at compile-time; at least, not without the same signature.
16:36:57	*	Jesin quit (Quit: Leaving)
16:37:42	disruptek	we really need more edit api for macros.
16:37:46	disruptek	this is ridiculous.
16:38:59	FromDiscord	<Recruit_main707> I was looking at the wiki, and openarray doesnt seem to need the toOpenArray function, is that outdated?
16:42:06	FromDiscord	<Doof Doof> Yo . . . so regarding macros, can we create some like in Haxe where people will convert assets, like images, into statically accessible names in the code?
16:42:11	FromDiscord	<Doof Doof> something like*
16:42:18	FromDiscord	<lqdev> sure
16:42:21	FromDiscord	<Doof Doof> Sweet
16:42:21	FromDiscord	<lqdev> check out `slurp()`
16:42:23	disruptek	shashlick has a package to do this already.
16:42:33	disruptek	!repo author:genotrance assets
16:42:34	disbot	no results 😢
16:42:39	disruptek	!repo author:genotrance embed
16:42:40	disbot	no results 😢
16:42:42	disruptek	!repo author:genotrance
16:42:43	disbot	no results 😢
16:42:51	disruptek	dafuq
16:42:54	FromDiscord	<Doof Doof> I'll find it!
16:43:15	FromDiscord	<Doof Doof> Thanks nimlings - is that the right name?
16:43:16	FromDiscord	<lqdev> I just use `slurp("my-sprite.png")` to embed stuff into my executables, then just load the resources into memory using appropriate APIs at runtime
16:43:30	FromDiscord	<lqdev> @Doof Doof we don't really have a name so it's fine :)
16:43:31	FromGitter	<sealmove> guys I have a problem which I can't solve for months... I have a NimNode of arbitary Nim expression and I want to infer its type. Can I use the compiler to do this somehow, or do I have to write a proc that iterates the node and does type algebra to figure it out?
16:44:00	FromDiscord	<lqdev> sealmove: pass the node to a macro which accepts a `typed` parameter, and in that macro use `getType`
16:47:48	FromGitter	<sealmove> signature is `macro infer(e: typed): NimNode` ?
16:48:38	FromDiscord	<lqdev> not really
16:48:50	FromDiscord	<lqdev> the thing is, you need to generate a call to that macro and pass your NimNode to `e`
16:48:57	FromDiscord	<lqdev> so you may need to pass in extra params
16:49:31	FromDiscord	<lqdev> it's kind of difficult to explain
16:49:34	FromDiscord	<lqdev> hold on a second
16:49:42	FromGitter	<sealmove> yes, people tried to explain this to me before :S
16:50:56	FromGitter	<sealmove> so I need a macro which generates a macro call?
16:51:07	FromDiscord	<Vindaar> yes
16:51:32	disruptek	i need to do something similar.
16:52:13	FromDiscord	<Vindaar> and trust me, you will probably hit many walls
16:53:48	FromGitter	<sealmove> yes, it's scary
16:54:08	FromDiscord	<Vindaar> things like: https://github.com/nim-lang/Nim/issues/13913
16:54:10	disbot	➥ Cannot hand multiple procs to macro using `varargs[typed]` ; snippet at 12https://play.nim-lang.org/#ix=2s5l
16:54:23	*	tane joined #nim
16:54:35	FromDiscord	<lqdev> sealmove: here's an example https://play.nim-lang.org/#ix=2s5m
16:54:52	FromDiscord	<lqdev> macros are a little half-baked.
16:55:41	FromDiscord	<Vindaar> and the infamous https://github.com/nim-lang/RFCs/issues/44
16:55:42	disbot	➥ [RFC] Working with types in macro is difficult.
16:58:02	*	Jesin joined #nim
16:58:09	*	marnix joined #nim
16:58:22	FromGitter	<sealmove> hmmmm
16:59:11	FromDiscord	<Vindaar> don't get discouraged. one can usually make it work, but it can be frustrating
16:59:13	FromGitter	<sealmove> lqdev: so you bindsymed echo to see the type in screen? I need it as string or NimNode in my program to use in some codegen
17:02:56	FromDiscord	<lqdev> well, there isn't really a way to get it back at this point
17:03:04	FromDiscord	<lqdev> you need to do any further work in the other macro
17:03:31	FromDiscord	<lqdev> I really wish Nim had a proc like `resolveIdent` that would turn idents into syms, but I can only wish…
17:03:37	FromDiscord	<lqdev> or open an RFC.
17:04:23	FromGitter	<sealmove> wow so it's not possible after all
17:04:23	*	leorize quit (Ping timeout: 240 seconds)
17:06:12	FromDiscord	<Vindaar> @lqdev well, it technically has. `bindSym` with the experimental "dynamicBindSym" it does just that
17:06:13	FromDiscord	<Vindaar> https://nim-lang.github.io/Nim/macros.html#bindSym%2C%2CBindSymRule
17:06:37	FromDiscord	<lqdev> no it doesn't. it binds the sym in the macro's scope
17:06:54	FromDiscord	<lqdev> but what is really needed is something that binds a sym from the callsite's scope
17:07:09	*	leorize joined #nim
17:08:03	FromDiscord	<Vindaar> wait, in the callsite's scope it's not really a problem though?
17:08:16	FromDiscord	<Vindaar> then you are in compiled nim land
17:08:45	FromDiscord	<Vindaar> and you can just construct some `typeof` call or whatever
17:09:16	FromDiscord	<Vindaar> but probably I misunderstand you
17:09:20	FromDiscord	<lqdev> but you can't get the type of an ident.
17:09:23	FromDiscord	<lqdev> you need a sym.
17:09:54	FromDiscord	<lqdev> if you're parsing a DSL that refers to some outside data, you need to defer to another macro. there is no way to turn untyped→typed.
17:10:28	FromGitter	<sealmove> So basically you need a full-fledged lookup table for your variables I guess
17:10:50	FromGitter	<sealmove> At least to track the types
17:10:52	leorize[m]	@lqdev bindSym with open scope doesn't work?
17:11:13	FromDiscord	<lqdev> with open scope? what does that mean?
17:11:30	*	marnix quit (Ping timeout: 260 seconds)
17:11:34	leorize[m]	bindSym take an optional parameter
17:11:50	leorize[m]	that specify the binding scope
17:12:51	leorize[m]	iirc brOpen should be the same as template's "mixin"
17:13:27	FromDiscord	<lqdev> ohhhhhh
17:13:35	*	marnix joined #nim
17:13:35	*	marnix quit (Read error: Connection reset by peer)
17:13:37	FromDiscord	<Vindaar> I mean: this works "just fine": https://play.nim-lang.org/#ix=2s5v
17:13:50	FromDiscord	<lqdev> sealmove: you're saved
17:18:31	shashlick	Doof Doof - nimdeps
17:19:48	FromGitter	<sealmove> Ok but an expression can have many things. It can have identifiers or literals.
17:21:23	FromDiscord	<Vindaar> exactly, and that's where things become problematic (even if dynamic bindSym wasn't experimental)
17:23:09	FromDiscord	<Vindaar> that's why I do all sorts of BS trying to figure out some type info in ggplotnim, here for instance https://github.com/Vindaar/ggplotnim/blob/master/src/ggplotnim/dataframe/arraymancer_backend.nim#L1008-L1075
17:23:48	FromDiscord	<Vindaar> and that already includes a 2 step macro starting at untyped ⇒ typed
17:26:56	*	NimBot joined #nim
17:27:03	FromGitter	<sealmove> :/
17:27:18	FromGitter	<raydf> Hello everyone
17:27:41	FromDiscord	<Vindaar> hey @raydf!
17:28:55	*	marnix joined #nim
17:29:43	FromGitter	<raydf> is there any example on creating custom blocks for wrapping statements with closures objects? Pseudo example of the usage i'll like: ⏎ ⏎ ```withConnection: ⏎ resp "#S" % [connection] ``` ⏎ ⏎ i want this code to handle the opening and closing of a db connection [https://gitter.im/nim-lang/Nim?at=5f148307724012319bb7f44e]
17:32:38	FromDiscord	<Vindaar> you can use a template like so: https://play.nim-lang.org/#ix=2s5B
17:36:09	FromGitter	<sealmove> Is it possible to create/manually constructed type NimNodes?
17:36:22	FromGitter	<sealmove> typed*
17:38:11	FromDiscord	<lqdev> nope, you can only get them via some user input or by using bindSym
17:38:25	FromDiscord	<lqdev> but you can create symbols using genSym
17:46:44	FromGitter	<raydf> thx @Vindaar, i was looking for the pragmas
17:52:51	FromGitter	<raydf> I'm having trouble with the template it opens the connection but the connection.close() is unreachable
17:54:50	FromDiscord	<Vindaar> can you show your code?
18:05:39	FromGitter	<raydf> of course
18:07:36	FromGitter	<raydf> https://play.nim-lang.org/
18:08:17	FromGitter	<raydf> https://play.nim-lang.org/#ix=2s5M
18:11:13	FromDiscord	<Doof Doof> anyone know why nim wouldn't be able to link to a dll but another language can to the same exact one? The dll is in the same folder, and in the parent folder, of the nim built exe
18:12:07	Prestige	@Vindaar thanks for helping with macros yesterday, I feel like I understand them now. Wrote a few more today
18:12:10	FromDiscord	<Doof Doof> Os is Win10
18:12:27	FromGitter	<deech> Is there a way to get Nim doc generation to list types and procs in a module in alphabetical order instead of the order in which they appear in the module? It would make things easier to find.
18:12:52	Zevv	+1
18:13:47	FromDiscord	<Vindaar> @Avahe you're welcome!
18:16:51	FromDiscord	<Vindaar> @raydf: I can't compile your code, but because of a different issue with `jesterwithplugins`
18:17:39	FromDiscord	<Vindaar> in principle this should work though (unless the macros around that tamper with the code generated by the template!)
18:19:24	FromGitter	<raydf> i tried with jester and is giving me the same error. I'll keep trying to understand how the jester macros exit before calling the disconnect code. Again, thx for the help
18:23:01	FromDiscord	<dom96> Hey guys, I'm working on a IO game called Stardust and would love you to give it a whirl. Give it a go here in your favourite browser: https://stardust.dev/play. And let me know what you think 🙂
18:26:13	FromDiscord	<Doof Doof> Still having trouble with dlls . . .
18:26:15	FromGitter	<raydf> @Vindaar found a workaround using a return value and then calling the resp macro. That macro breaks and leave the rest of the code without executing
18:26:32	FromDiscord	<Doof Doof> @dom96 I killed you lol, I was `nimling` - but again . . .too easy to be swept from below
18:27:05	leorize	@dom96 well you gotta fix the "can be killed during spawn" :P
18:27:09	FromDiscord	<Doof Doof> I keep seeing these funny usernames. . . "Implement in rust" Fix nim plz" - "async is broken". What's all this about lol?
18:27:31	shashlick	@leorize - did you get to see that plugins crash with arc
18:28:03	FromDiscord	<dom96> leorize: pff, it's easy to escape spawn killing
18:28:07	leorize	shashlick: yea, it's complicated due to the lack of stacktraces, even in gdb
18:28:30	shashlick	ok
18:28:32	FromDiscord	<dom96> Yeah, who are these people asking to fix nimble and async? 😛
18:28:34	leorize	it's kinda laggy on my firefox, but then I've like canvasblocker running :P
18:28:57	Zevv	"Could not find server / Request failed / Okay"
18:29:06	FromDiscord	<dom96> Yeah, performance on FF sucks.
18:29:08	FromDiscord	<dom96> Chrome is much better
18:29:14	shashlick	dom96: pinged you on discord about https://github.com/dom96/choosenim/pull/201#issuecomment-639774677
18:29:15	disbot	➥ Fix #199 - missing dlls
18:29:36	FromDiscord	<dom96> (If you're on Linux with FF then it's really bad, or at least it was for some)
18:30:11	FromDiscord	<dom96> Zevv: hm, browser? region?
18:30:19	Zevv	ff nl
18:30:41	FromDiscord	<dom96> Can't connect at all?
18:30:48	Zevv	Cross0origin request blocked, related?
18:30:52	leorize	how does the exploding thingy work?
18:30:54	FromDiscord	<dom96> anything interesting in the console?
18:31:00	Zevv	"Reason: CORS header accessc-ontrol-allow-origin missing
18:31:07	FromDiscord	<dom96> leorize: click to burst and then aim at others
18:31:30	FromDiscord	<Doof Doof> lol wth, why "whenfastasync"?
18:31:44	FromDiscord	<Doof Doof> all these names are making me think something wrong haha
18:31:45	FromDiscord	<dom96> someone trolling lol
18:32:18	FromDiscord	<dom96> Someone from Argentina 🤔
18:32:56	FromDiscord	<dom96> > "Reason: CORS header accessc-ontrol-allow-origin missing↵@Zevv[IRC]#0000 can you see which server it's trying to connect to?
18:32:58	*	krux02 joined #nim
18:33:25	Zevv	"Cross-Origin Request Blocked: The Same Origin Policy disallows reading the remote resource at https://stardust.dev/route/server. (Reason: CORS header ‘Access-Control-Allow-Origin’ missing).
18:34:11	Zevv	~Cross-Origin Request Blocked: The Same Origin Policy disallows reading the remote resource at https://stardust.dev/route/server. (Reason: CORS header ‘Access-Control-Allow-Origin’ missing).
18:34:12	disbot	no footnotes for `Cross-Origin`. 🙁
18:34:14	FromDiscord	<dom96> But how can that be? It's on the same domain?
18:34:18	Zevv	{"url":"wss://fra1.stardust.dev/game/ws"}
18:34:44	FromDiscord	<dom96> Oh, so you're getting a response
18:34:55	FromDiscord	<dom96> Maybe it cannot connect to FRA1
18:35:10	FromDiscord	<dom96> Does it work in other browsers for you?
18:35:14	Zevv	lemme check
18:35:25	Zevv	nope
18:35:25	FromDiscord	<Doof Doof> so if I have a dll in the same directory as my exe? I'm thinking about just using a full path instead of a dll name. It's getting ridiculous.
18:35:46	FromDiscord	<Doof Doof> directory as my exe . . . why can't it load? *
18:36:27	leorize	@dom96 I got a ton of mass ejected (as a ring) and become super slow after, why's that?
18:36:43	FromDiscord	<dom96> Zevv: can you check the network tab and see what isn't being loaded?
18:37:04	FromDiscord	<dom96> leorize: that's by design, just to add a little more nuance to the gameplay 🙂
18:37:11	Zevv	uhm, how do I see what's *not* being loaded?
18:37:15	*	couven92 joined #nim
18:37:19	*	fredrikhr quit (Disconnected by services)
18:37:26	*	couven92 is now known as fredrikhr
18:37:38	Zevv	I get the startdust.dev/server request with the wws response, and that's it
18:37:41	FromDiscord	<Doof Doof> Yeah . . . even a fullpath isn't connecting the dll. What the heck
18:37:46	*	couven92 joined #nim
18:37:55	leorize	wrong architecture?
18:38:35	FromDiscord	<Doof Doof> From the nim code or the dll? I am using the exact same dll in another language. . .
18:38:39	FromDiscord	<dom96> Zevv: here is what it looks like for me when I refresh then click "Play Online" https://media.discordapp.net/attachments/371759389889003532/734479328796082236/unknown.png
18:38:54	Zevv	not at all like that
18:38:58	Zevv	just one request
18:38:59	Zevv	and it stalls
18:39:24	FromDiscord	<dom96> Can you show me a screenshot?
18:40:56	FromDiscord	<Doof Doof> @leorize
18:41:16	FromDiscord	<Doof Doof> (edit) '@leorize' => '@leorize-'
18:41:30	leorize	nim --version
18:41:36	leorize	what architecture is nim compiled for?
18:41:58	FromDiscord	<Doof Doof> i386
18:42:08	leorize	and what architecture is your dll?
18:43:27	FromDiscord	<Doof Doof> I'll have to check. I didn't create the dll. But if it helps - I can't load DLL's created by the nim compiler either.
18:43:42	leorize	hmm, how are you doing it?
18:44:24	FromDiscord	<Doof Doof> For the one's that were created by the nim compiler, I was testing out hot-reloading. It just wouldn't load the dll no matter where I got it from.
18:44:44	FromDiscord	<Doof Doof> I'm currently just loading the dll I'm playing with right now using raylib forever, and it looks like it's using "dynlib"
18:45:04	FromDiscord	<Doof Doof> `# {.pragma: RLAPI, cdecl, discardable, dynlib: "raylib" & LEXT.}` where lext resolves to .dll
18:50:42	*	marnix quit (Ping timeout: 256 seconds)
18:54:25	FromDiscord	<Doof Doof> also - building the exe works fine. it's literally just the dll load :/
18:54:33	FromDiscord	<Doof Doof> I don't get why one lang works and another doesn't
18:55:13	FromDiscord	<Doof Doof> only thing I've had success with so far in Nim is just some console printing and input
18:59:31	*	narimiran joined #nim
19:05:48	*	Tlanger quit (Ping timeout: 258 seconds)
19:11:11	*	marnix joined #nim
19:13:46	*	fredrikhr quit (Ping timeout: 256 seconds)
19:19:36	*	fredrikhr joined #nim
19:23:45	leorize	hmmm this is weird
19:23:56	leorize	my guess is that you're using choosenim?
19:25:12	FromDiscord	<Doof Doof> choosenim?
19:25:59	FromDiscord	<Doof Doof> no I used the manual installation
19:29:06	*	hoijui joined #nim
19:43:54	leorize	manual? that's weird, isn't that zip file 64bit...
19:44:50	leorize	https://github.com/alaviss/nightlies/releases <- grab the latest nightlies here and test with that
19:45:17	*	ForumUpdaterBot quit (Remote host closed the connection)
19:45:20	leorize	nim devel have enhanced error message for dll loads, so maybe it will reveal some more issues
19:45:24	*	ForumUpdaterBot joined #nim
19:51:51	FromDiscord	<dom96> > my guess is that you're using choosenim?↵Wouldn’t happen if choosenim was used 😉
19:52:55	FromDiscord	<Doof Doof> hahaha ^
19:53:01	FromDiscord	<Doof Doof> Okay I'll try both those in a bit
19:55:46	leorize	:P was guessing purely on how choosenim prefers 32bit binaries
20:04:33	shashlick	it doesn't anymore
20:04:42	shashlick	if no gcc is present, it picks 64-bit
20:04:49	shashlick	if gcc is present, it uses it whatever it is
20:05:11	*	Vladar quit (Quit: Leaving)
20:06:49	FromDiscord	<Doof Doof> Okay I ran choosenim, what am I supposed to expect? It says I have stable but nothing really happened . . .
20:07:44	*	fanta1 quit (Quit: fanta1)
20:11:38	shashlick	trying to figure out what's your issue
20:14:23	shashlick	so your nim is not 64-bit - presume gcc -v says the same
20:14:33	shashlick	but you don't know what arch the dll is?
20:15:00	*	mal`` quit (Quit: Leaving)
20:15:30	*	mal`` joined #nim
20:15:40	FromDiscord	<Doof Doof> the arch should be 64
20:15:53	FromDiscord	<Doof Doof> That's what the language/compiler I was using prior was
20:16:00	FromDiscord	<Doof Doof> so it worked with that architecture
20:16:33	shashlick	if nim -v and gcc -v are not 64-bit then it won't work with this dll
20:16:52	shashlick	http://www.dependencywalker.com/ is very useful for arch and what the dll depends on
20:17:02	shashlick	but ya, did you install gcc separately?
20:17:47	FromDiscord	<Doof Doof> yeahhh I did download the 32 bit somehow
20:18:05	FromDiscord	<Doof Doof> not sure how I goofed that one up
20:18:40	shashlick	no problem, just get rid of nim and gcc, then run choosenim again - it will download the 64-bit gcc
20:18:52	shashlick	alternative is to setup gcc yourself
20:19:18	leorize	shashlick: does choosenim detect windows native architecture or it just guess?
20:19:56	FromDiscord	<slymilano> @dom96 Araq asked me to open a bug report here: https://github.com/nim-lang/Nim/issues/15019
20:19:57	disbot	➥ asyncCheck works fine on Linux and Windows - not on Mac. ; snippet at 12https://play.nim-lang.org/#ix=2s6h
20:19:58	shashlick	it detects cpu arch
20:20:41	shashlick	https://github.com/dom96/choosenim/blob/a7b9d24849255509b57a2f4fb7d97d6e79a860e8/src/choosenimpkg/cliparams.nim#L106
20:21:31	FromDiscord	<Doof Doof> i think that actually worked
20:21:32	FromDiscord	<Doof Doof> nice
20:21:34	FromDiscord	<Doof Doof> let me check
20:21:53	FromDiscord	<Doof Doof> dooooooope
20:22:03	FromDiscord	<Doof Doof> I see a bouncing ball on my screen and I've never been happier
20:22:22	FromDiscord	<Doof Doof> not an exaggeration at all
20:24:04	FromDiscord	<Doof Doof> Sorry for the last few bits of spam there . . . but just wanted to thank everyone for their help!! Now it's time to start a potential journey in Nim . . .
20:24:43	leorize	shashlick: there's also windows' GetNativeSystemInfo if you want to be 100% correct :P
20:25:14	leorize	maybe we can switch to that once microsoft debut their arm windows 10 X tablet
20:25:39	shashlick	nift
20:28:47	FromDiscord	<dom96> Wow. I was joking about choosenim. Nice that it actually resolved your issue. Maybe now less folks will be negative about it 🙂
20:29:17	skrylar[m]	wasn't aware people were
20:34:32	shashlick	@dom96 - that PR is good to go now - ptal - CI is failing for windows since 0.6.0 is broken, not the build that is being tested
20:35:33	*	fredrikhr quit (Disconnected by services)
20:35:39	*	couven92 is now known as fredrikhr
20:41:21	*	idf31 joined #nim
20:42:47	*	couven92 joined #nim
20:49:15	*	couven92 quit (Quit: Client disconnecting)
20:58:07	*	nikita` quit (Quit: leaving)
20:59:10	*	marnix quit (Ping timeout: 256 seconds)
20:59:55	*	marnix joined #nim
21:01:24	FromDiscord	<treeform> @dom96 I only use choosenim now.
21:11:10	*	marnix quit (Read error: Connection reset by peer)
21:12:45	*	Zectbumo joined #nim
21:18:08	FromDiscord	<Avatarfighter> choosenim best nim
21:19:05	*	tane quit (Quit: Leaving)
21:22:24	*	narimiran quit (Quit: leaving)
21:40:04	idf31	sup bros, some weeks ago I made a game in Nim using the Nico framework for the 8x8 Game Jam 2(were sprites had to be maximum 8x8, fantasy console-esque framework was perfect) and was wondering if you would be interested
21:40:34	idf31	the code could surely be improved and I still have to add more levels, but i will do those after the jam rating period ends for fairness i guess
21:41:29	FromDiscord	<Elegant Beef> What do you mean by interested
21:41:39	FromDiscord	<Elegant Beef> Also there is a nim gamedev IRC btw
21:42:00	idf31	oh thanks
21:42:12	idf31	by interested i meant if i can post it or something
21:42:37	FromDiscord	<Rika> i mean why not
21:43:01	FromDiscord	<Elegant Beef> #nim-gamedev for the irc
21:43:14	FromDiscord	<Elegant Beef> Yea i mean show off what you made/make is fine here
21:46:36	idf31	https://idf04.itch.io/door-o-bot
21:48:32	FromDiscord	<Elegant Beef> Need to not have the music/movement sfx play so loud/often 😄
21:48:49	idf31	yeah sorry about that
21:53:27	FromDiscord	<Doof Doof> Please correct me if I'm wrong - but the whole "style insensitive" typing is not applied to the reserved words?
21:56:16	FromDiscord	<Yardanico> It is applied to all symbols
21:56:25	FromDiscord	<Yardanico> Including keywords
21:56:38	FromDiscord	<Yardanico> There are no exceptions where it's not applied really
21:56:51	FromDiscord	<Elegant Beef> Yea it'd be weird to get exceptions for it
21:56:58	FromDiscord	<Doof Doof> huh . . .
21:56:59	FromDiscord	<Doof Doof> well
21:57:17	FromDiscord	<Yardanico> notin is same as not_in or notIn for example
21:57:20	FromDiscord	<Doof Doof> I tried capitalizing `Import` and `While` but both resulted in `undeclared identifier`
21:57:27	FromDiscord	<Elegant Beef> cause first letter matters
21:57:31	FromDiscord	<Elegant Beef> try `iMpoRt`
21:57:32	FromDiscord	<Yardanico> no, you didn't read it carefully enough :P
21:57:42	FromDiscord	<Yardanico> First letter in Nim is case sensitive
21:57:44	FromDiscord	<Yardanico> For all symbols
21:58:03	FromDiscord	<Doof Doof> Darn. I was hoping I could write it like I used to write in my favorite language
21:58:06	FromDiscord	<Yardanico> That's mainly so you can have types and variables with the same name
21:58:15	FromDiscord	<Yardanico> @Doof Doof no one stops you from making templates
21:58:20	FromDiscord	<Elegant Beef> just imagine it as removing `_` and doing `toLower()` on all symbols
21:58:25	FromDiscord	<Yardanico> Or a macro to transform "your favourite language" to Nim
21:58:25	FromDiscord	<Doof Doof> oh well ¯\_(ツ)_/¯
21:58:33	FromDiscord	<Yardanico> What's that "favourite language" by the way?
21:58:34	FromDiscord	<Doof Doof> @Yardanico all good! I don't need them to be cap'd
21:58:34	FromDiscord	<Elegant Beef> C#?
21:58:38	FromDiscord	<Doof Doof> No, Monkey2
21:59:08	FromDiscord	<Elegant Beef> statically typed lua basically?
21:59:25	FromDiscord	<Doof Doof> It's closer related to Basic dialects
21:59:37	FromDiscord	<Elegant Beef> And yard you though i wrote code weirdly https://media.discordapp.net/attachments/371759389889003532/734529903327707156/HzxfyQ.png
21:59:43	FromDiscord	<Elegant Beef> (edit) 'though' => 'thought'
21:59:53	FromDiscord	<Doof Doof> Has operator overloading, classes, inheritance, static extensions, and a dead language
21:59:54	FromDiscord	<Doof Doof> etc
22:00:03	FromDiscord	<Elegant Beef> `( a:int, b:string )`
22:00:04	FromDiscord	<Doof Doof> Pretty damn good FFI too
22:00:21	FromDiscord	<Elegant Beef> Yea just get use to camelCasing 😛
22:00:39	FromDiscord	<Doof Doof> I'm already used to it - I use Haxe pretty frequently
22:00:44	FromDiscord	<Doof Doof> :P
22:01:04	FromDiscord	<Rika> rest in peace
22:01:33	FromDiscord	<Elegant Beef> I hope it doesnt if that's the typical convention that language needed to die
22:01:52	FromDiscord	<Doof Doof> Well let me see . . .
22:02:04	FromDiscord	<Doof Doof> It let you FFI with C++ or Java pretty easily
22:02:15	FromDiscord	<Doof Doof> Java for android development
22:02:18	FromDiscord	<Elegant Beef> i mean nim has C/C++ FFI
22:02:25	FromDiscord	<Doof Doof> It also had objective C interfacing
22:02:32	FromDiscord	<Doof Doof> I think it was swift actually . . . let me check
22:02:44	FromDiscord	<Doof Doof> I could be wrong ¯\_(ツ)_/¯
22:02:56	FromDiscord	<Doof Doof> I haven't used this in over a year for basically anything
22:03:12	*	hoijui quit (Ping timeout: 260 seconds)
22:03:22	FromDiscord	<Elegant Beef> Well we also do have↵https://github.com/yglukhov/jnim
22:03:35	FromDiscord	<Elegant Beef> But out of the box nim does do C/C++/js/objc FFI
22:03:54	FromDiscord	<Doof Doof> Nim's got better ffi I think
22:04:15	FromDiscord	<Doof Doof> well - almost certainly. Monkey2 was not really more than just for making games to be fair
22:07:44	FromDiscord	<Doof Doof> Anyways - all good, I was just double checking earlier about the syntax stuff and I'm still happy anyhow
22:11:37	*	solitudesf quit (Ping timeout: 264 seconds)
22:16:10	*	idf31 quit (Ping timeout: 256 seconds)
22:20:23	*	lainon joined #nim
22:41:23	*	aenesidemus joined #nim
22:46:21	FromDiscord	<Varriount> I can't recall anyone actually using the Objective C interface recently.
22:46:40	FromDiscord	<Varriount> It's one of those "it worked in the past, presumably it still works" features.
22:48:39	FromDiscord	<XxDiCaprioxX> I mean if Nim compiles to C it should work
22:48:48	FromDiscord	<Yardanico> Objective C is a separate backend
22:48:51	FromDiscord	<Yardanico> Well, not take
22:48:54	FromDiscord	<Yardanico> Really*
22:49:09	FromDiscord	<Yardanico> Objective C "backend" is a part of the C backend
22:49:12	FromDiscord	<XxDiCaprioxX> But does it really matter if it is the same language
22:49:17	FromDiscord	<Yardanico> It's not
22:49:25	FromDiscord	<Yardanico> Objective C has differences from C
22:50:20	FromDiscord	<Yardanico> Ah well, it seems to be a superset
22:50:47	FromDiscord	<Yardanico> But Nim obj c backend is for allowing you to interface with objective c libraries
23:00:13	*	ofelas joined #nim
23:10:54	*	ofelas quit (Ping timeout: 256 seconds)
23:12:02	*	lainon quit (Ping timeout: 260 seconds)
23:15:08	*	krux02_ joined #nim
23:19:07	*	krux02 quit (Ping timeout: 265 seconds)
23:20:52	*	ofelas joined #nim
23:33:37	*	ofelas quit (Ping timeout: 260 seconds)
23:39:26	*	lainon joined #nim
23:43:19	*	lainon quit (Client Quit)
23:44:44	*	fredrikhr quit (Ping timeout: 265 seconds)
23:48:55	*	ofelas joined #nim
23:54:10	*	oddp quit (Quit: quit)